Home > Research > Publications & Outputs > Clustering piecewise stationary processes

Electronic data

  • nstclust

    Accepted author manuscript, 276 KB, PDF document

    Available under license: CC BY-NC: Creative Commons Attribution-NonCommercial 4.0 International License

View graph of relations

Clustering piecewise stationary processes

Research output: Contribution in Book/Report/Proceedings - With ISBN/ISSNConference contribution/Paper

Forthcoming

Standard

Clustering piecewise stationary processes. / Khaleghi, Azadeh; Ryabko, Daniil.

IEEE International Symposium on Information Theory. 2020.

Research output: Contribution in Book/Report/Proceedings - With ISBN/ISSNConference contribution/Paper

Harvard

Khaleghi, A & Ryabko, D 2020, Clustering piecewise stationary processes. in IEEE International Symposium on Information Theory.

APA

Khaleghi, A., & Ryabko, D. (Accepted/In press). Clustering piecewise stationary processes. In IEEE International Symposium on Information Theory

Vancouver

Khaleghi A, Ryabko D. Clustering piecewise stationary processes. In IEEE International Symposium on Information Theory. 2020

Author

Khaleghi, Azadeh ; Ryabko, Daniil. / Clustering piecewise stationary processes. IEEE International Symposium on Information Theory. 2020.

Bibtex

@inproceedings{707f4c3d426a42ffb86a07c9f9fb1405,
title = "Clustering piecewise stationary processes",
abstract = "The problem of time-series clustering is considered in the case where each data-point is a sample generated by a piecewise stationary process. While stationary processes comprise one of the most general classes of processes in nonparametric statistics, and in particular, allow for arbitrary long-range dependencies, their key assumption of stationarity remains restrictive for some applications.We address this shortcoming by considering piecewise stationary processes, studied here for the first time in the context of clustering. It turns out that this problem allows for a rather natural definition of consistency of clustering algorithms. Efficient algorithms are proposed which are shown to be asymptotically consistent without any additional assumptions beyond piecewise stationarity. The theoretical results are complemented with experimental evaluations.",
author = "Azadeh Khaleghi and Daniil Ryabko",
year = "2020",
month = mar
day = "31",
language = "English",
booktitle = "IEEE International Symposium on Information Theory",

}

RIS

TY - GEN

T1 - Clustering piecewise stationary processes

AU - Khaleghi, Azadeh

AU - Ryabko, Daniil

PY - 2020/3/31

Y1 - 2020/3/31

N2 - The problem of time-series clustering is considered in the case where each data-point is a sample generated by a piecewise stationary process. While stationary processes comprise one of the most general classes of processes in nonparametric statistics, and in particular, allow for arbitrary long-range dependencies, their key assumption of stationarity remains restrictive for some applications.We address this shortcoming by considering piecewise stationary processes, studied here for the first time in the context of clustering. It turns out that this problem allows for a rather natural definition of consistency of clustering algorithms. Efficient algorithms are proposed which are shown to be asymptotically consistent without any additional assumptions beyond piecewise stationarity. The theoretical results are complemented with experimental evaluations.

AB - The problem of time-series clustering is considered in the case where each data-point is a sample generated by a piecewise stationary process. While stationary processes comprise one of the most general classes of processes in nonparametric statistics, and in particular, allow for arbitrary long-range dependencies, their key assumption of stationarity remains restrictive for some applications.We address this shortcoming by considering piecewise stationary processes, studied here for the first time in the context of clustering. It turns out that this problem allows for a rather natural definition of consistency of clustering algorithms. Efficient algorithms are proposed which are shown to be asymptotically consistent without any additional assumptions beyond piecewise stationarity. The theoretical results are complemented with experimental evaluations.

M3 - Conference contribution/Paper

BT - IEEE International Symposium on Information Theory

ER -